← Back to Publications List

AI-Driven Predictive Models for Flood Forecasting and Disaster Management in Bangladesh

Students & Supervisors

Student Authors
Mubassir Islam Jimel
Bachelor of Science in Computer Science & Engineering, FST
Samiha Sultana
Bachelor of Science in Computer Science & Engineering, FST
Md. Abdur Rahman
Bachelor of Science in Computer Science & Engineering, FST
Shohag Rana
Bachelor of Science in Computer Science & Engineering, FST
Maruf Ahammed
Bachelor of Science in Computer Science & Engineering, FST
Supervisors
Md. Mortuza Ahmmed
Associate Professor, Faculty, FST

Abstract

Bangladesh is a country affected by frequent floods, which are caused by complicated hydro-climatic and fluvial-climatic relations. Climate change has also made these risks severe through the increase in the variability of rainfall, the rise in the intensity of river flow regimes, and the prevalence of extreme events. The objective of this research is to examine the trends of three significant indicators, such as Rainfall Index, River Discharge Index, and Flood Impact Score, in the period between 1995 and 2024, with the aim of drawing some conclusions that will reinforce AI-based forecasting and disaster management models. The secondary data on trend analysis has been processed and visualized using the methodology of excel based graphs of secondary data gathered through different sources. The results indicate that the River Discharge Index showed serious change when it reached its peak in 2002 and sharply declined in 2004, and on the other hand, the Rainfall Index did not show any significant variation throughout the period of study. The scores of Flood Impact have been increasing significantly in 2013 and 2018, which is a sign of high-impact years of floods. The indicators show weak correlations, thus showing that there is no single factor that drives the severity of floods in Bangladesh, but rather a combination of several factors. Such findings underscore the necessity to construct multi-input nonlinear AI models comprising the integration of various indicators into a dependable early warning mechanism. Some of the strategies to integrate AI should be hybrid machine learning models, feature engineering, seasonal cycles, a combination of sequence models, and socioeconomic vulnerability levels to enhance predictability. By using the power of AI to put together multiple sources of data in flood prone areas, future systems would be able to provide long-term and workable insight into the problems.

Keywords

Flood forecasting; AI-driven modeling; Socio economic impact; Disaster risk management.

Publication Details

  • Type of Publication:
  • Conference Name: 2nd International Conference on Frontiers in Science: Innovation & Technology for Greener Industry (2nd ICFS:ITGI)
  • Date of Conference: 15/01/2026 - 15/01/2026
  • Venue: BUET, Dhaka-1000
  • Organizer: Faculty of Science, BUET, Dhaka 1000, Bangladesh
View on Publisher